Using data from SEER databases, Howlader and colleagues recently reported that incidence-based mortality from NSCLC in men decreased from 6.3% annually (95% CI, 3.4 to 9.0) from 2013 through 2016, in contrast to the period between 2006 and 2013, when it decreased by 3.1% annually (95% CI:2.8 to 3.3); In women, incidencebased mortality decreased slowly, by 2.3% annually from 2006 through 2014, and then at a faster rate of 5.9% annually from 2014 through 2016. Lung cancer–specific survival also improved substantially. These improvements were seen for all races and ethnic groups.1
The investigators reported incidencebased mortality, which looks at lung cancer incidence relative to lung cancer mortality. By using SEER data, they were able to evaluate NSCLC and SCLC separately, and were able to determine that the decrease in incidence-based mortality in NSCLC was probably due to changes in treatment, whereas mortality in SCLC was almost entirely due to declining incidence, with no improvement in survival. This method of estimating mortality by comparing it to incidence has been used to assess the effect of prevention vs. intervention in other cancer sites.1 Prevention decreases incidence but not mortality, whereas an intervention will decrease mortality without a corresponding decrease in incidence.
Why would a paper that showed a modest improvement in incidence-based mortality of roughly 3% over a 10 year span be considered of enough significance to be published in the prestigious New England Journal of Medicine? Several reasons, perhaps the most important of which is the inflection point —the acceleration in the rate of decline in incidence-based mortality occurred in the years 2013 to 2014. And what was happening during this period of time? The advent of targeted therapy. Erlotinib and gefitinib were approved for the first-line treatment of patients with stage IV, EGFR-mutation positive metastatic lung cancer in 2013 and 2015, respectively, and crizotinib was approved for the treatment of first-line metastatic ALK-positive NSCLC in 2013. Other papers had reported a decrease in lung cancer mortality, but without breaking the results by subtypes, they were unable to determine that these results were most likely due to improvements in treatment.2
Another reason that these findings are important is because they show a significant change in the slope of the lung cancer mortality curve. For decades, despite the development of more effective and less toxic chemotherapy, second-line chemotherapy, new combinations, maintenance therapy, and anti-angiogenic agents, improvements in lung cancer mortality were few, small, and far between. The 1995 Non-small Cell Lung Cancer Collaborative Group published a landmark meta-analysis showing the benefit of cisplatin-based chemotherapy: 5% overall 1-year survival of patients with untreated NSCLC compared to 15% overall 1-year survival for patients treated with cisplatin based chemotherapy.3 In the Howlader paper, 2-year lung cancer specific survival was 35% in men diagnosed in 2014, and 44% in 2014 for women.
What had not yet happened in 2013? The approval of checkpoint inhibitors, which occurred in 2015. Checkpoint inhibitors are estimated to have been given to a majority of patients with stage III and IV NSCLC, with clinical trials ongoing to determine their effectiveness in the adjuvant setting. Lung cancer screening with low-dose CT, approved in 2014, is underutilized with only 4.4 % of eligible persons screened in 2015.4 Second and third-generation TKIs and identification of additional targetable mutations were also in the offing. We look forward to seeing further improvements in the incidence, incidence-based mortality, and survival of patients with lung cancer in the next analysis of these data.